This proposal is a collaborative effort by a team of five motor systems groups at five institutions seeking to probe the mechanisms that underlie the brain's capacity for learning a new motor skill. The common thread for all groups is to focus on changes that occur within the primary motor cortex as a new skill is acquired. Changes in motor cortex will be characterized in relationship to critical input areas including premotor and parietal cortex and the role of subcortical circuits in learning will also be modeled. Both immediate and long-lasting changes of motor cortex representation will be investigated using a synthesis of molecular, cellular, systems and computational level of analysis. Project 1: Dr. Peter Strick (at the University of Pittsburgh) will combine flavoprotein optical imaging and single unit recording in monkey to characterize changes of activity in premotor and motor cortex as animals learn sequential behavior. Flavoprotein imaging allows for long-term cortical mapping over at least two years time, making it possible to look at dynamic alterations of cortical neuronal activity throughout the training period. Project 2: Dr. Scott Grafton (at University of California, Santa Barbara) will use functional MRI and transcranial magnetic stimulation in humans to study the neural substrates for off-line consolidation of three types of newly acquired motor skills: sequencing, visuomotor and dynamic adaptation. The tasks are similar to those in the other projects allowing for translation between monkey and human studies. Project 3: Dr. Emilio Bizzi (at MIT) will collaborate with experts in nanotechnology and conducting polymers at MIT to develop a new type of electrode based on fine wires of conducting polymers. With this he will perform chronic recordings of primary motor neurons in primates learning to move in novel dynamics. Project 4: Dr. James Houk (at Northwestern University) and Dr. Andrew Barto (at the University of Massachusetts at Amherst) will develop computational models of learning that are integral to the other projects of this PPG. These models will be used to explore critical behavioral and representational issues. Core A: Support for program administration and videoconferencing to achieve PPG integration. In addition, the core will support 3 satellite conferences on motor learning and mechanisms of cortical reorganization that are relevant for translational research. These interrelated projects, focusing on single anatomical substrate and common set of learning behaviors, should provide an integration of methodologies across multiple levels of analysis that are far beyond those achievable if each project were pursued separately. The collaborative effort can be expected to significantly advance our knowledge about mechanisms that support motor cortex plasticity.